Social Security provision in Mexico is not universal. There is a mix of state and federal programs. In June 2007, the Mexican federal government announced a social security program for the elderly 70 years or over in towns with fewer than 2,500 inhabitants. This was extended to towns with fewer than 10,000 inhabitants in February 2008. The State of Yucatan is implementing a social security program for people 75 or over for towns above 10,000 inhabitants. Thus the Yucatan program does not have overlap with the Federal program. People over age 75 get a pension of about $50 per month plus a food basket, worth about $7. People between ages 65 and 75 only get a food basket. The program is phased in over a five year period, to be completed by the end of 2011. The state government of Yucatan has asked RAND to design, conduct, and evaluate this social policy intervention. Several program parameters are still subject to change, e.g. how often to give money (once a month or more often), whether to apply means testing (a desire by the state), content of the food basket, the age 75 cut-off, etc. We propose a research study that aims to inform the optimal design of the new social security program and that evaluates its effects on the health and well-being of the elderly recipients. We will implement an experimental design with control and treatment groups, and frequently measure a large number of outcomes, including health (measured by biomarkers and self-reports), income, wealth, consumption, economic activities, mental well-being, social contacts, etc. Moreover we are planning several experiments to determine optimal program parameters (e.g. the content of the food basket, or the mode of payment). This will be the first evaluation of a non-contributory pension program with an experimental design including treatment and control groups measuring the effects mainly on health outcomes of the oldest old. The project is partly financed by the State of Yucatan and by existing resources at RAND. We are asking support for the effort of the researchers in designing the evaluation and experiments, as well as preliminary analyses of the outcomes. PUBLIC HEALTH RELEVANCE: The State of Yucatan is implementing a social security program for people 75 or over for towns above 10,000 inhabitants. We will design and evaluate the program and experiment with various program parameters, using a randomized design.